From f7d7d632a2cc3bc2305da6511d4cf2606db42f45 Mon Sep 17 00:00:00 2001 From: Ralf Corsepius Date: Thu, 8 Sep 2005 12:47:42 +0000 Subject: Experimental commit. --- doc/FAQ/endoftime.texi | 135 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 135 insertions(+) create mode 100644 doc/FAQ/endoftime.texi (limited to 'doc') diff --git a/doc/FAQ/endoftime.texi b/doc/FAQ/endoftime.texi new file mode 100644 index 0000000000..4664235baf --- /dev/null +++ b/doc/FAQ/endoftime.texi @@ -0,0 +1,135 @@ +@c +@c COPYRIGHT (c) 1988-2002. +@c On-Line Applications Research Corporation (OAR). +@c All rights reserved. +@c +@c $Id$ +@c + + +@node Date/Time Issues in Systems Using RTEMS, Hardware Issues, , Top + +@chapter Date/Time Issues in Systems Using RTEMS +@ifinfo +@menu +* Hardware Issues:: +* RTEMS Specific Issues:: +* Language Specific Issues:: +* Date/Time Conclusion:: +@end menu +@end ifinfo + +This section provides technical information regarding +date/time representation issues and RTEMS. The Y2K problem has +lead numerous people to ask these questions. The answer to +these questions are actually more complicated than most +people asking the question expect. RTEMS supports multiple +standards and each of these standards has its own epoch and +time representation. These standards include both programming +API and programming language standards. + +In addition to the issues inside RTEMS +itself, there is the complicating factor that the Board +Support Package or application itself may interface with hardware +or software that has its own set of date/time representation +issues. + +In conclusion, viewing date/time representation as "the Y2K problem" +is very short-sighted. Date/time representation should be viewed as +a systems level issue for the system you are building. Each software +and hardware component in the system as well as the systems being +connected to is a factor in the equation. + + +@node Hardware Issues, RTEMS Specific Issues, Date/Time Issues in Systems Using RTEMS, Date/Time Issues in Systems Using RTEMS + +@section Hardware Issues + +Numerous Real-Time Clock (RTC) controllers provide only a two-digit +Binary Coded Decimal (BCD) representation for the current year. Without +software correction, these chips are a classic example of the Y2K problem. +When the RTC rolls the year register over from 99 to 00, the device +has no idea whether the year is 1900 or 2000. It is the responsibility +of the device driver to recognize this condition and correct for it. +The most common technique used is to assume that all years prior +to either the existence of the board or RTEMS are past 2000. The +starting year (epoch) for RTEMS is 1988. Thus, + +@itemize @bullet +@item Chip year values 88-99 are interpreted as years 1988-2002. +@item Chip year values 00-87 are interpreted as years 2000-2087. +@end itemize + +Using this technique, a RTC using a +two-digit BCD representation of the current year will overflow on +January 1, 2088. + + +@node RTEMS Specific Issues, Language Specific Issues, Hardware Issues, Date/Time Issues in Systems Using RTEMS + +@section RTEMS Specific Issues + +Internally, RTEMS uses an unsigned thirty-two bit integer to represent the +number of seconds since midnight January 1, 1988. This counter will +overflow on February 5, 2124. + +The time/date services in the Classic API will overflow when the +RTEMS internal date/time representation overflows. + +The POSIX API uses the type @i{time_t} to represent the number of +seconds since January 1, 1970. Many traditional UNIX systems as +well as RTEMS define @i{time_t} as a signed thirty-two bit integer. +This representation overflows on January 18, 2038. The solution +usually proposed is to define @i{time_t} as a sixty-four bit +integer. This solution is appropriate for for UNIX workstations +as many of them already support sixty-four bit integers natively. +At this time, this imposes a burden on embedded systems which are +still primarily using processors with native integers of thirty-two +bits or less. + + +@node Language Specific Issues, Date/Time Conclusion, RTEMS Specific Issues, Date/Time Issues in Systems Using RTEMS + +@section Language Specific Issues + +The Ada95 Language Reference Manual requires that the @i{Ada.Calendar} +package support years through the year 2099. However, just as the +hardware is layered on top of hardware and may inherit its limits, +the Ada tasking and run-time support is layered on top of an operating +system. Thus, if the operating system or underlying hardware fail +to correctly report dates after 2099, then it is possible for the +@i{Ada.Calendar} package to fail prior to 2099. + + +@node Date/Time Conclusion, , Language Specific Issues, Date/Time Issues in Systems Using RTEMS + +@section Date/Time Conclusion + +Each embedded system could be impacted by a variety of date/time +representation issues. Even whether a particular date/time +representation issue impacts a system is questionable. A system +using only the RTEMS Classic API is not impacted by the +date/time representation issues in POSIX. A system not using +date/time at all is not impacted by any of these issues. Also +the planned end of life for a system may make these issues +moot. + +The following is a timeline of the date/time representation +issues presented in this section: + +@itemize @bullet + +@item 2000 - Two BCD Digit Real-Time Clock Rollover + +@item 2038 - POSIX @i{time_t} Rollover + +@item 2088 - Correction for Two BCD Digit Real-Time Clock Rollover + +@item 2099 - Ada95 @i{Ada.Calendar} Rollover + +@item 2124 - RTEMS Internal Seconds Counter Rollover + +@end itemize + + + -- cgit v1.2.3